candidate-sourcing

Warn

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: MEDIUMDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [DYNAMIC_EXECUTION]: The skill utilizes a JavaScript execution tool to orchestrate complex API workflows, fan out requests, and process data. It also includes a Python script template for managing candidate tracking in CSV format.
  • [INDIRECT_PROMPT_INJECTION]: The pipeline ingests candidate data from varied external sources like GitHub and academic papers to generate personalized outreach. This creates a surface for indirect prompt injection as external content is processed and incorporated into agent outputs.
  • Ingestion points: Data is fetched via search and enrichment tools from LinkedIn, GitHub, and web searches (SKILL.md).
  • Boundary markers: There are no specific delimiters or instructions to ignore embedded commands in the fetched content.
  • Capability inventory: The skill possesses file system write capabilities for CSV logging and the ability to create Gmail drafts (SKILL.md).
  • Sanitization: The instructions do not specify sanitization or escaping of external content before its use in email generation.
  • [EXTERNAL_DOWNLOADS]: The skill fetches candidate information and metadata from GitHub's official API and commit pages to verify identity and find contact information.
  • [COMMAND_EXECUTION]: The skill provides a Python script for logging candidate details to the local file system.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 27, 2026, 01:00 PM
Security Audit — agent-trust-hub — candidate-sourcing